Thomas Turnbull

Thomas Turnbull is completing his PhD at the School of Geography and the Environment (SOGE) at the University of Oxford. He earlier completed a degree in history and worked in the environmental sector. At Oxford Thomas has lectured on digital research methods and controversy mapping, and at University College London (UCL), alongside Professor Andrew Barry, he has taught classes on mapping environmental controversies online. More recently, he was been awarded The Oxford Research in the Humanities (TORCH) Environmental Humanities award to support an international conference «Controlling Environments». Thomas interests lie in the political and scientific history of energy, cybernetics, intellectual history, and historical geography. Most recent publications: «Scientific Visualisation in Practice: Replicating Experiments at Scale». In: Leonardo, 48 (1), pp. 72-73; «Review of Carbon Democracy: Political Power in the Age of Oil». In: Area, 46 (1), pp. 115-117.



The Computation of Energy Conservation

My thesis documents the history of energy conservation, meant in terms of energy resource conservation rather than in the strict thermodynamic sense, though the two are related. The project spans from 1800’s to the present day, and describes how the science and politics of energy conservation were manifest in Victorian Britain, during the Cold War, the energy crises of the 1970s, and electricity market privatisation in the 1990s. My central argument is that for energy conservation to occur the energy system had to be reconfigured as a computational device. As such, the history of energy conservation must be situated in reference to the wider history of computation, systems theory and cybernetics. In combining the history of energy with the history of computing, I hope to make a novel argument about the relationship between computerisation and resource economics. 

Of direct relevance to the MECS project, my intention is to more fully explore the idea that energy conservation is a computational science. This thesis is partly based on a close analysis of the work of RAND Corporation mathematician Fred Roberts (1943--) who in 1970, in response to America’s growing demand for energy, developed a calculative method for forecasting energy demand via graph theory.  Using signed diagraphs, Roberts suggested, not only could energy demand be dynamically forecasted, but it could also be programmed. If the energy system were configured so that it had an odd number of negative informational feedbacks, the system would automatically counteract any increase (‘deviation’) in energy demand. Such a structure was intended to create systemic memory which, Roberts argued, could automate the regulation of societal energy demand, thereby conserving energy which otherwise would have been reallocated elsewhere in the system.

At MECS I hope to more fully address the computational history of energy conservation, and the simulation of energy demand, and to reflect upon this engagement with the support of the MECS community. Before working at RAND, Roberts’ was taught by computer scientist Dana Scott (1932-) whose work on finite automata was closely linked to Alan Turing’s work on ‘universal machines’ and Warren McCulloch and Walter Pitt’s work on ‘neural nets’. Given the significance of these figures in the history of computing, I wish to explore the relation between Roberts’ work and the wider intellectual history of automation and simulation. Alongside this genealogical work, I hope to be able to reconstruct aspects of Roberts’ energy demand forecasting model, and to reflect on this simulation to reconsider the relation between computation and the political economy of energy resources.